CVE Database

llama.cpp builds b7492 through the latest b9060 contains a use-after-free vulnerability in llama-server affecting six tokenization endpoints (/tokenize, /detokenize, /infill, /apply-template, /rerank, and /anthropic/count_tokens) that bypass the task queue and access ctx_server.vocab directly on HTTP worker threads. Attackers can exploit a time-of-check-time-of-use race condition where the main thread destroys and frees vocab after the synchronization lock is released but before the handler finishes using it, causing a crash or potential code execution when --sleep-idle-seconds is configured.

A race condition in JCacheCodeDataProvider allows an attacker to redeem a single authorization code multiple times via concurrent requests, resulting in the issuance of multiple distinct, valid access tokens. Users are recommended to upgrade to versions 4.2.3, 4.1.8 or 3.6.12, which fix this issue.

PraisonAI is a multi-agent teams system. In versions prior to 1.6.58, the web_crawl tool performs its SSRF check only on the initially supplied URL, allowing the protection to be bypassed so the tool connects to attacker-chosen internal destinations. The check resolves the hostname once with socket.gethostbyname and rejects private/loopback/link-local results, but then passes the URL to a fetcher using httpx.Client(follow_redirects=True) (or urllib.request.urlopen when httpx is absent, which also follows redirects) that re-resolves the hostname at connect time with no further validation. This validate-here/fetch-there gap is exploitable through both HTTP redirects and DNS rebinding. If an attacker can influence URLs passed to web_crawl(), directly or through an agent/tool workflow, they can cause the PraisonAI host to fetch loopback, private-network, or cloud metadata endpoints reachable from that host, with the response body returned in the web_crawl() result. This issue has been fixed in version 1.6.58.

Memos' webhook dispatch function safeDialContext (internal/webhook/webhook.go) resolves the target hostname via net.DefaultResolver.LookupHost and validates the resulting IPs against reserved ranges, but then dials net.JoinHostPort(host, port) using the original hostname rather than the already-validated IP address.

Misskey is an open source, federated social media platform. Versions 12.37.0 and later, but prior to 2026.5.4, are vulnerable to timing attacks during JSON-LD signature validation and the compaction process. Because the JSON-LD parsing context is not shared between signature verification and subsequent processing, the application may trust information that should not be trusted, resulting in a time-of-check to time-of-use (TOCTOU) flaw. This allows an attacker to have fraudulent activities accepted as valid, leading to a loss of integrity. This issue has been fixed in version 2026.5.4.

Netty is an asynchronous, event-driven network application framework. Prior to versions 4.1.136.Final and 4.2.16.Final, the OcspServerCertificateValidator forwards the SslHandshakeCompletionEvent before the asynchronous OCSP validation completes. This allows the client's downstream handlers to send sensitive application data (e.g., HTTP requests) to a revoked server before the channel is closed by the OCSP check. n io.netty.handler.ssl.ocsp.OcspServerCertificateValidator#userEventTriggered, when an SslHandshakeCompletionEvent is received, the validator immediately calls ctx.fireUserEventTriggered(evt). It then initiates an asynchronous OCSP query using OcspClient.query. Because the handshake completion event is forwarded immediately, downstream handlers in the client's pipeline are notified that the TLS handshake is successful. They may then begin reading and processing incoming application data or sending outgoing data. If the OCSP response later indicates the server's certificate is REVOKED, the validator closes the channel, but by this time, the client may have already leaked sensitive data to a revoked server or processed malicious responses from it. This issue has been fixed in versions 4.1.136.Final and 4.2.16.Final.

When grant-copy operations are processed, the respective grant may or may not already be in use by another operation (a mapping or another copy). For all copy operations the referenced guest frame is looked up. When another operation is already active for the grant (the grant is "pinned"), what is being supplied back to actually carry out permission checks and copy operation may not be consistent: The permission check may be carried out on a page different from the one involved in the copy.

In the Linux kernel, the following vulnerability has been resolved: KVM: x86: Ensure vendor's exit handler runs before fastpath userspace exits Move the handling of fastpath userspace exits into vendor code to ensure KVM runs vendor specific operations that need to run before userspace gains control of the vCPU. E.g. for VMX (and soon to be for SVM as well), KVM needs to flush the PML buffer prior to exiting to userspace, otherwise any memory written by the final KVM_RUN might never be flagged as dirty. Note, waiting to snapshot CR0 and CR3 until svm_handle_exit() is flawed in general, as that risks consuming stale state in a fastpath handler. That will be addressed in a future change.

n8n before 1.123.64, 2.29.8, and 2.30.1 contains a TOCTOU race condition in the Git node's clone operation that allows authenticated users to bypass path restrictions by swapping a directory for a symlink after the path is validated but before the clone runs. This lets an attacker plant a crafted repository in the community node directory, which n8n loads as a custom node on the next restart, executing arbitrary JavaScript on the server. Both self-hosted and cloud instances are affected.

In the Linux kernel, the following vulnerability has been resolved: rbd: eliminate a race in lock_dwork draining on unmap Given how rbd_lock_add_request() and rbd_img_exclusive_lock() are written, lock_dwork may be (re)queued more than it's actually needed: for example in case a new I/O request comes in while we are in the middle of rbd_acquire_lock() on behalf of another I/O request. This is expected and with rbd_release_lock() preemptively canceling lock_dwork is benign under normal operation. A more problematic example is maybe_kick_acquire(): if (have_requests || delayed_work_pending(&rbd_dev->lock_dwork)) { dout("%s rbd_dev %p kicking lock_dwork\n", __func__, rbd_dev); mod_delayed_work(rbd_dev->task_wq, &rbd_dev->lock_dwork, 0); } It's not unrealistic for lock_dwork to get canceled right after delayed_work_pending() returns true and for mod_delayed_work() to requeue it right there anyway. This is a classic TOCTOU race. When it comes to unmapping the image, there is an implicit assumption of no self-initiated exclusive lock activity past the point of return from rbd_dev_image_unlock() which unlocks the lock if it happens to be held. This unlock is assumed to be final and lock_dwork (as well as all other exclusive lock tasks, really) isn't expected to get queued again. However, lock_dwork is canceled only in cancel_tasks_sync() (i.e. later in the unmap sequence) and on top of that the cancellation can get in effect nullified by maybe_kick_acquire(). This may result in rbd_acquire_lock() executing after rbd_dev_device_release() and rbd_dev_image_release() run and free and/or reset a bunch of things. One of the possible failure modes then is a violated rbd_assert(rbd_image_format_valid(rbd_dev->image_format)); in rbd_dev_header_info() which is called via rbd_dev_refresh() from rbd_post_acquire_action(). Redo exclusive lock task draining to provide saner semantics and try to meet the assumptions around rbd_dev_image_unlock().